In this report, we describe the three-dimensional skin reconstruct model which mimics human skin in architecture and composition. Melanocyte physiology, melanoma progression and the fate of dermal stem cells have been investigated using the skin reconstruct model. The model is also useful as a preclinical tool for drug assessment.
Most in vitro studies in experimental skin biology have been done in 2-dimensional (2D) monocultures, while accumulating evidence suggests that cells behave differently when they are grown within a 3D extra-cellular matrix and also interact with other cells (1-5). Mouse models have been broadly utilized to study tissue morphogenesis in vivo. However mouse and human skin have significant differences in cellular architecture and physiology, which makes it difficult to extrapolate mouse studies to humans. Since melanocytes in mouse skin are mostly localized in hair follicles, they have distinct biological properties from those of humans, which locate primarily at the basal layer of the epidermis. The recent development of 3D human skin reconstruct models has enabled the field to investigate cell-matrix and cell-cell interactions between different cell types. The reconstructs consist of a “dermis” with fibroblasts embedded in a collagen I matrix, an “epidermis”, which is comprised of stratified, differentiated keratinocytes and a functional basement membrane, which separates epidermis from dermis. Collagen provides scaffolding, nutrient delivery, and potential for cell-to-cell interaction. The 3D skin models incorporating melanocytic cells recapitulate natural features of melanocyte homeostasis and melanoma progression in human skin. As in vivo, melanocytes in reconstructed skin are localized at the basement membrane interspersed with basal layer keratinocytes. Melanoma cells exhibit the same characteristics reflecting the original tumor stage (RGP, VGP and metastatic melanoma cells) in vivo. Recently, dermal stem cells have been identified in the human dermis (6). These multi-potent stem cells can migrate to the epidermis and differentiate to melanocytes.
1. The Three-Dimensional Human Skin Reconstruct Model:
2. Representative Results:
Skin reconstructs with normal melanocytes and dermal stem cells
The skin reconstructs are cultured in a special 6-well tray with inserts. The dermal compartment is cultured in DMEM with 10% FBS for the first four days (Fig. 1A). The skin reconstruct medium I is used when keratinocytes are seeded with melanocytes or melanoma cells (Fig. 1B). The epidermis is exposed to the air at day 9 and this allows keratinocytes to differentiate (Fig. 1C). At day 18, the epidermis of skin reconstructs is composed of stratified keratinocyte layers. The undifferentiated basal layer and sequentially differentiated layers are vertically oriented (Fig. 1D). Staining the section of reconstructs with the melanocytic marker S100 shows that melanocytes are aligned in the basal layer of the epidermis and communicate with multiple keratinocytes through dendrite extensions (Fig. 1E). The dermal compartment contains fibroblasts embedded in a collagen type I matrix. Deposited collagen IV indicates the basement membrane, which separates the epidermis from the dermis (Fig. 1F). When dermal stem cells (labeled with GFP lentiviral vector) are embedded with fibroblasts in a collagen I matrix, they migrate to the epidermis and differentiate into melanocytes (1), (Fig. 2). Multiple layers of keratinocytes in the epidermis are developed.
Skin reconstructs of melanoma tumors
Clinicopathological studies indicate that melanomas progress in a stepwise manner: common acquired nevi, dysplastic nevi, RGP (radial growth phase) melanomas, VGP (vertical growth phase) melanomas and metastatic melanomas (7). Different stages of melanoma cell lines are morphologically similar to each other in 2-D culture (Fig. 3A-D) but when they are incorporated in skin reconstructs, the behavior of the cells reflects their in vivo characteristics. The location and the growth rate of normal melanocytes are tightly controlled in skin reconstructs (Fig. 3E). RGP primary melanomas WM35 proliferate predominantly in the epidermis (Fig. 3F), whereas VGP melanomas WM793 grow invasively into the dermis (Fig. 3G). Metastatic melanomas 1205Lu aggressively invade deep into the dermis (Fig. 3H).
Figure 1. Skin reconstructs with normal melanocytes. The gross appearance of the skin reconstructs is shown in A-C. A. Fibroblasts mixed with collagen are grown in DMEM containing 10% FBS and form dermal compartment. B. Keratinocytes and melanocytes are seeded on top of the dermis and grow in skin reconstruct medium. C. The epidermis is exposed to air at day 9. D. H&E-stained skin reconstruct presents the epidermis comprised vertically, oriented basal layer, and sequentially differentiated stratified cell layers. The dermis contains fibroblasts embedded in a collagen type I matrix. E. S100-positive melanocytes (black arrows) are aligned at the basement membrane and communicate with multiple keratinocytes. F. Collagen IV-staining indicates the basement membrane, which separates the epidermis from the dermis. All the stainings in D-F were performed on formalin-fixed, paraffin-embedded sections.
Figure 2. Dermal stem cells in skin reconstructs migrate to the epidermis and differentiate into melanocytes. At day 5 after seeding keratinocytes, single GFP-positive cells (green) start migrating out from spheres. The epidermis is still composed of a single layer. At day 8, a few cells reach the epidermis-dermis interface. By day 10, GFP-positive cells are tightly aligned at the basement membrane position. The migrated GFP-positive cells in the epidermis express melanocytic marker HMB45 (red, as indicated by white arrows). Nuclei are stained with DAPI (blue). The epidermis is developed as multiple layers. The basement membrane is indicated with white dotted lines.
Figure 3. Skin reconstructs of different stages of melanomas: A-D. Normal melanocytes and melanoma cells grown in 2D cultures. A. Melanocytes from foreskin. B. RGP WM35 cells. C. VGP WM793 cells. D. Metastatic melanoma 1205Lu cells. E. Normal melanocytes are located at the basement membrane. F. RGP melanoma WM35 cells grow as cell clusters in the epidermis. G. VGP melanoma WM793 cells invade into the dermis through basement membrane. H. Metastatic melanoma 1205Lu cells aggressively invade deep into dermis.
Troubleshooting
Problem | Troubleshooting |
Collagen mixture does not solidify | Collagen mixture color should be straw-yellow to pink, otherwise pH is wrong and collagen may not gel. If the color is bright yellow, more sodium bicarbonate should be added drop by drop. |
Collagen prematurely precipitates in the mixuture | Keep all components on ice until the collagen mixture is placed onto the insert |
Contracted collagen is not even (one side is thicker than the other side) | Calibrate the shelf of incubator |
The epidermis is formed with less than three keratinocyte layers. | Use undifferentiated keratinocytes at lower passages |
We have described generating 3D skin reconstructs with normal human melanocytes, dermal stem cells and melanoma cells. When grown in monolayer culture, melanocytic cells present a similar morphology (flattened, spindle or more dendritic-shaped) regardless of their origin of clinical stage. In contrast, 3D skin reconstructs recapitulate stage-specific properties of melanoma cells. Normal human melanocytes reside at the basement membranes between the epidermal and the dermal layers as single cells. RGP primary melanoma cells grow as small clusters along the basement membrane whereas more aggressive VGP melanoma cells grow as large clusters breaking through the basement membrane into the dermis. Metastatic melanoma cells grow in all directions and invade deep into the dermis as single cells or clusters. Quantitative analyses can be performed on this model by measuring the depth of invasion and the extent of proliferation. In addition to characterization of normal and malignant melanocytic cells, using the model we can directly induce the differentiation of dermal stem cells into matured melanocytes, which home to the basal layer of epidermis and establish bona fide communication with keratinocytes through upregulation of E-cadherin (6).
Using viral vectors, we are able to either activate or inactivate gene function for a better understanding of the dynamics and functional significance of genes expressed by each cell type in skin reconstructs, which promises to be an efficient model to study not only transformation mechanisms of melanocytes, but also progression of melanoma (8). Long-term observation of cell phenotypes has become possible through grafting of skin reconstructs to immunodeficient animals. Such a human skin-mouse chimera is an excellent research tool to study melanomagenesis and melanoma metastasis.
Skin reconstructs can also be a useful platform for drug assessment. Many drugs which eradicate cancer cells in 2D culture conditions often have little effect in experimental and clinical applications. As seen in the in vivo tumor, melanoma cells in 3D cultures are often resistant to the drugs which cells in 2D cultures respond to, suggesting that the microenvironment modulates signaling pathways in melanoma. For successful drug discovery, the 3D skin reconstruct model is an ideal preclinical tool to predict the effects of compounds in vivo (9,10).
In summary, skin reconstruct models bridge the gap between in vitro and in vivo studies. They will lead to a better understanding of which genes are involved in transformation and how stem cells contribute to that transformation.
The authors have nothing to disclose.
We thank the Wistar Institute Animal Facility, Microscopy Facility, Histotechnology Facility and Research Supply Center. This study was funded in part by grants from the National Institutes of Health CA 076674, CA 098101, CA 025874 and CA 10815.
Name of the reagent | Company | Catalogue number |
---|---|---|
10x EMEM | Bio Whittaker | 12-684F |
L-glutamine | Gibco | 25030-081 |
Fetal Bovine Serum | Hyclone | SH-30071.02 |
Bovine Tendon Acid-Extracted Collagen | Organogenesis | 200/50 |
Tissue Culture 6-well Trays with Inserts | Organogenesis | 9285 |
Keratinocyte Serum Free Medium | Gibco | 17005042 |
Dialyzed Fetal Calf Serum | Hyclone | SH-30079.03 |
recombinant Human Stem Cell Factor | Fitzgerald Industries | RDI-307-255X |
Basic FGF | Fitzgerald Industries | 30R-AF015 |
Endothelin-3, human | American Peptide Co. | 88-5-10 |
Calcium Chloride | Sigma | C-7902 |
DMEM | JRH biosciences | 56430-10L |
F12 (HAM’s) | Gibco | 11765-54 |
Hydrocortisone | Sigma | H-0888 |
Insulin, Transferrin, Ethanolamine, Selenium (ITES) | BioWhittaker | 17839Z |
O-Phosphorylethanolamine | Sigma | P0503 |
Adenine | Sigma | A9795 |
Progesterone | Sigma | P-8783 |
Triiodothyronine | Sigma | T-5516 |
Newborn calf serum | Hyclone | SH3011802 |
10% buffer formalin | Surgipath | 00600 |
Optimal cutting temperature freezing media (OCT) | Sakura | 4583 |
Multi cassettes | Surgipath | 02293-BX |
TBS biopsy papers | Triangle Biomedical Sciences | BP-B |
TBS biopsy papers | Triangle Biomedical Sciences | BP-B |
Disposable Base Molds | Fisher | 15-182-501C |
Compound benzoin tincture | Professional Disposables, Inc, | S42450 |
Alcohol Prep Swabs | CVS pharmacy | |
Silk Black Braided 5-0 | Ethicon | 682G |
Gauze sponges (sterile) | CVS pharmacy | |
Forceps | Roboz | RS5070 |
Iris scissors | Roboz | RS5913 |
Surgical blades | Feather | 2976 |
EZ anesthesia | Euthanex Corp. | |
SCID hairless outbred mouse | Charles river |